A critical issue regarding aging and cognition concerns the impact of strategic behavior as a cause of age differences in cognitive performance. The proposed research continues our research on the influence of strategic factors on age differences in skill acquisition. We have adopted the noun-pair learning (NP) task (Ackerman &Woltz, 1994) to assess how quickly individuals switch from a slower form of controlled processing (visual scanning of an array to find target word pairs) to more fluent performance based on memory-retrieval. A substantial proportion of older adults do not rely on retrieval from memory to make rapid judgments in the NP task, opting instead to stay with the slower but effective strategy of visual scanning. This persistent scanning behavior cannot be fully explained by deficient rates of mechanistic associative learning - the shift to retrieval is also influenced by age differences in strategy choice. Using trial-level strategy reports and recognition memory probes, we have shown that older adults avoid memory retrieval despite adequate noun-pair learning, slowing their skill acquisition. Our specific aims are to: (1) produce further evidence of older adults'retrieval avoidance with experimental context effects on rates of retrieval shift that are inconsistent with a pure associative learning deficit;(2) test the flexibility of initial response criteria and strategy preferences by manipulating instructional sets and reward contingencies;(3) check the generality of these effects by extending them to other skill acquisition tasks, and evaluating consistent conservative choice behavior across a variety of task domains, and (4) conduct extensive individual differences studies that evaluate the relationship of strategic choice to (a) background abilities, such as perceptual speed and associative memory;(b) personal characteristics, including memory self-concept, task mental model, and metacognitive monitoring ability;and (c) individual differences in online metacognition, including a fast feeling- of-knowing and updating of knowledge for latency differences across strategies. We evaluate whether older adults'retrieval avoidance can be overcome by experimental intervention.